[00:36:02] Newsletter 56: Proposed summary of the licensing discussion. Welcome, Mariya! Office hour on Monday at 19:00 UTC - https://meta.wikimedia.org/wiki/Abstract_Wikipedia/Updates/2021-12-16 [01:13:00] I tried my own summary. I hope those two together work well (re @mahir256: I see that an attempt at a vote summary has been provided on the page. Lookes pretty accurate, but perhaps some nuance is being lost?) [08:28:00] Well, it seems my entire argument was discarded, so it feels like my contributions to the discussion were a total waste of time [08:29:06] I feel like the whole project just got a lot less interesting for me :( [08:31:07] I absolutely hate having to deal with licenses, which is why I like contributing to CC0 projects like wikidata and why I try to release things I make as CC0 too, so I basically won't be able to use anything from wikifunctions or abstract wikipedia? [08:31:32] and if I can't use it, why bother contributing? [08:32:14] and since licenses like cc by-sa are viral, we'll be stuck with it forever [08:33:55] maybe lucas should keep not wikifunctions around after all :P [08:34:03] maybe lucas should keep not wikilambda around after all :P [08:47:26] I hope all these non-english-speaking non-programmers that we'd like to attract have a better grasp of licenses and when and how to correctly attribute anything they reuse than I do, and I really hope we don't attract any copyright trolls like commons has [08:57:17] I disagree with you about the license for content, but you and I are in the same boat because my comments about the license for implementation were ignored, too 🤷🏻‍♂️ (re @Jan_ainali: Well, it seems my entire argument was discarded, so it feels like my contributions to the discussion were a total waste of time) [08:57:50] I haven't seen a good explanation why Apache is good. [08:58:25] It's fine to disagree, that's why we need to have a discussion in the first place :) (re @amire80: I disagree with you about the license for content, but you and I are in the same boat because my comments about the license for implementation were ignored, too 🤷🏻‍♂️) [09:18:09] I'm not convinced by arguments about forcing people to use the same license, there are far too many people willing to redo things from scratch in order to do things their own way (including me) and big companies certainly have enough money to do that, but I don't really understand why apache would be the best choice either, it seems more like a poor compromise between the people like me who want things to be unrestricted and [09:19:00] I'm not convinced by arguments about forcing people to use the same license, there are far too many people willing to redo things from scratch in order to do things their own way (including me) and big companies certainly have enough money to do that, but I don't really understand why apache would be the best choice either, it seems more like a poor compromise between the people like me who want things to be unrestricted and [09:20:39] +1 to concerns re: lack of license choice on initial launch, despite my preference for copyleft at a certain high-level in both content and implementation and not-so below it [09:40:45] There will be many situations where one will be comfortable with a default rendering of a fact rather than some display customization thereof, and there will also be situations where the (reduced) complexity of a function does not grant us enough of a competitive advantage to require any license protections that GPL offers [11:43:29] it seems a rather good compromise has been reached: everyone's unhappy about it :P [13:18:10] References to actual code/content examples as you and I floated got thrown out as well during the discussion, which is as much a shame (re @Jan_ainali: Well, it seems my entire argument was discarded, so it feels like my contributions to the discussion were a total waste of time) [13:19:59] +1 to this in the absence of license choice (re @Nikki: maybe lucas should keep not wikilambda around after all :P) [13:20:55] and I wonder if I should relicense Ninai/Udiron as GPLv3 out of spite 😜 [18:57:07] Thank you for your answers here. I want to reassure you that we have read all the comments in detail, and thought about what was said . The brief summary is not meant to be a comprehensive mirror of the discussion - I will add that to the text, to make that explicit. I understand that in a case like this we cannot make everyone happy. So, Jan, Mahir, Nikki, Amir, others - [18:58:29] I am sorry that this proposed decision is disappointing, but I hope you agree the result is consistent with the result of the discussion, even if the summary is too brief. [18:59:08] By the way, you're more than welcome to suggest changes to the summary text, to reflect the differences better [19:12:48] I haven't read the discussion, so I can't say whether or not the result is consistent with it, but if it is then disappointing is an understatement [19:19:32] not only because I had naively assumed it would be cc0 by default like wikidata, but also because if most people who are involved *want* to place restrictions on things, then most people's values are different from mine [19:26:36] I can see your point, but we're not discussing simple data [19:27:11] (I explained why I think CC BY-SA was better on wiki, and I don't have the time now to explain myself better) [19:27:29] (I explained why I think CC BY-SA was better in the on-wiki discussion, and I don't have the time now to explain myself better) [19:27:39] (sorry for this) [19:28:10] I can say though that it isn't just because I want to restrict things, not even out of the blue [19:28:40] The summary does not reflect the discussion, at least on the topic of the license for the function implementation, which is the part that interests me. It mostly reflects the initial proposal. [19:28:52] I gave it a thought, I can see the point for CC0 and nevertheless I don't think it'd have been the choice I'd have made [19:31:01] The license for the implementation interests me the most because that's the part where the most significant effort of the volunteer Wikimedian developers will be invested. (re @amire80: The summary does not reflect the discussion, at least on the topic of the license for the function implementation, which is the part that interests me. It mostly reflects the initial proposal.) [19:39:31] +1 to this (re @amire80: The summary does not reflect the discussion, at least on the topic of the license for the function implementation, which is the part that interests me. It mostly reflects the initial proposal.) [21:14:39] Apache2 license choice is neither viral (CC-BY & GPL in other projects) nor unencumbered (CC0). Possibly more confusing MediaWiki is GPLv2 which is not considered to be compatible with Apache2 by the FSF. [21:17:28] Is this without 'or at your option any later version'? If so, why? (re @wmtelegram_bot: Apache2 license choice is neither viral (CC-BY & GPL in other projects) nor unencumbered (CC0). Possibly more confusing MediaWiki is GPLv2 which is not considered to be compatible with Apache2 by the FSF.) [21:18:23] @mahir256: MediaWiki is GPLv2+, but in practice the + part is for forks and the official codebase in very much GPLv2. [21:20:48] Sure, but are there parts which are v2 without the plus? If not, where's the harm in relicensing the software powering Wikifunctions (and not that for other Wikimedia projects) under v3? (re @wmtelegram_bot: @mahir256: MediaWiki is GPLv2+, but in practice the + part is for forks and the official codebase in very much GPLv2.) [21:22:32] (maybe you answered that and I read "GPLv2" as "plus unspecified", for which I apologize) [21:23:04] *"plus unspecified" as opposed to "with plus" or "without plus" [21:23:24] I'm not sure that you would be able to find an attorney representing the Foundation who would agree that such dual licensing is possible without opt-in approval from everyone who has ever contributed code to MediaWiki core. [21:25:31] The extent of what 'or at your option any later version' really means in a legal sense is undefined as far as I understand GPL enforcement. MediaWiki does not use a CLA to reassign copyright from the code contributor to the Foundation so relicensing requires consent. [21:29:45] If Abstract Wikipedia is to be pioneering in a number of areas, then why not in attempting to get that phrase defined in the legal arena? ;-) [22:57:46] Which phrase? [22:58:21] "or at your option any later version' [22:59:14] (the phrase Bryan said was undefined in a legal sense) (re @vrandecic: Which phrase?) [23:34:42] I don't hope we'll go there [23:35:46] In what sense isn't Apache not unencumbered? [23:42:21] @vrandecic: I was thinking about https://www.apache.org/licenses/LICENSE-2.0#redistribution and the attribution notice requirement for derivative works vs CC0 [23:44:58] My musing about GPLv2 and Apache2 incompatibly is just musing. I don't believe anyone would assert that the GPLv2 MediaWiki cannot execute Apache2 code, just that Apache2 code cannot be redistributed as part of a GPLv2 licensed project. [23:45:44] Although I guess this might have implications for any bootstrap code shipped in the extension itself. [23:47:00] That can probably be fixed by licensing the extension under GPLv3 which is deemed compatible with Apache2 by both FSF and ASF [23:48:52] FOSS licensing is a tar pit [23:49:50] can I get an amen for that? (re @wmtelegram_bot: FOSS licensing is a tar pit) [23:50:30] I’ve assumed that the bootstrap code is effectively CC0 [23:50:45] given that the extension, of its own accord, added it to a wiki which was already declared CC0 at that point [23:54:33] The structured data is CC0, and remains so. [23:55:47] I don't think there are any bootstrap code implementations that are part of the shipped extension. [23:56:05] hm, that’s possible [23:56:22] CC0 for code and the general stance of the Foundation on FOSS licensing runs into https://opensource.org/faq#cc-zero -- Generally the Foundation requires an OSI approved license and the OSI does not consider CC0 to be free. [23:56:52] CC0 cannot be considered a free license, being a copyright waiver [23:57:07] Oh, wow, I wasn't aware of that wrinkle